From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S266100AbUAGSik (ORCPT ); Wed, 7 Jan 2004 13:38:40 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266177AbUAGSik (ORCPT ); Wed, 7 Jan 2004 13:38:40 -0500 Received: from fw.osdl.org ([65.172.181.6]:41423 "EHLO mail.osdl.org") by vger.kernel.org with ESMTP id S266100AbUAGSii (ORCPT ); Wed, 7 Jan 2004 13:38:38 -0500 Date: Wed, 7 Jan 2004 10:38:31 -0800 (PST) From: Linus Torvalds To: Greg KH cc: Andrey Borzenkov , linux-hotplug-devel@lists.sourceforge.net, linux-kernel@vger.kernel.org Subject: Re: removable media revalidation - udev vs. devfs or static /dev In-Reply-To: <20040103055847.GC5306@kroah.com> Message-ID: References: <200401012333.04930.arvidjaar@mail.ru> <20040103055847.GC5306@kroah.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 2 Jan 2004, Greg KH wrote: > > Doesn't the kernel always create the main block device for this device? > If so, udev will catch that. But udev should probably also create all the sub-nodes if it doesn't already. And it really has to create _all_ of them, exactly because there's no way to know ahead-of-time which of them will be available. Then, user space can just access "/dev/sda1" or whatever, and the act of accessing it will force the re-scan. Linus